Citation

Oftedal S, Bell KL, Davies PSW, Ware RS, Boyd RN. Med. Sci. Sports Exerc. 2014; 46(9): 1808-1815.

Affiliation

1Queensland Cerebral Palsy and Rehabilitation Research Centre, School of Medicine, The University of Queensland, Brisbane, AUSTRALIA; 2Children's Nutrition Research Centre, Queensland Children's Medical Research Institute, The University of Queensland, Brisbane, AUSTRALIA; 3Queensland Children's Medical Research Institute, The University of Queensland, Brisbane, AUSTRALIA; and 4School of Population Health, The University of Queensland, Brisbane, AUSTRALIA.

Copyright

(Copyright © 2014, Lippincott Williams and Wilkins)

DOI

10.1249/MSS.0000000000000299

PMID

25134003

Abstract

AIM: The purpose of this study was to validate uni- and triaxial ActiGraph cut points for sedentary time in toddlers with cerebral palsy (CP) and typically developing children (TDC).

METHODS: Children (n = 103, 61 boys, mean age = 2 yr, SD = 6 months, range = 1 yr 6 months-3 yr) were divided into calibration (n = 65) and validation (n = 38) samples with separate analyses for TDC (n = 28) and ambulant (Gross Motor Function Classification System I-III, n = 51) and nonambulant (Gross Motor Function Classification System IV-V, n = 25) children with CP. An ActiGraph was worn during a videotaped assessment. Behavior was coded as sedentary or nonsedentary. Receiver operating characteristic-area under the curve analysis determined the classification accuracy of accelerometer data. Predictive validity was determined using the Bland-Altman analysis.

RESULTS: Classification accuracy for uniaxial data was fair for the ambulatory CP and TDC group but poor for the nonambulatory CP group. Triaxial data showed good classification accuracy for all groups. The uniaxial ambulatory CP and TDC cut points significantly overestimated sedentary time (bias = -10.5%, 95% limits of agreement [LoA] = -30.2% to 9.1%; bias = -17.3%, 95% LoA = -44.3% to 8.3%). The triaxial ambulatory and nonambulatory CP and TDC cut points provided accurate group-level measures of sedentary time (bias = -1.5%, 95% LoA = -20% to 16.8%; bias = 2.1%, 95% LoA = -17.3% to 21.5%; bias = -5.1%, 95% LoA = -27.5% to 16.1%).

CONCLUSION: Triaxial accelerometers provide useful group-level measures of sedentary time in children with CP across the spectrum of functional abilities and TDC. Uniaxial cut points are not recommended.

Language: en